Process variation refers to the differences in how a process is executed from one instance to another (for example from the Happy path and/or the Most Common Path to the other way in performing the process).
For example, if a process involves filling out a form, there might be variations in the time taken to complete the form, the number of errors made, or the number of steps involved. Process variation can be caused by a number of factors, such as differences in the input data, variations in the skills and knowledge of the process actors, or changes in the process itself.
Process variation can be a problem because it can lead to inefficiencies, errors, and inconsistencies in the output of the process. This is the reason why analyzing the process variation is a very important part of the Process Analysis, and as it is a high level analysis it’s recommended to do it first. Moreover, in the context of Lean Six Sigma as reducing process variation is an important goal, such analysis can help to improve the quality and efficiency of the process.
At this stage and to ensure the Process conformity we won’t check one by one all the Process variations. What is interesting is to take a look at the Process variations coverage (How many Process flows follow this path, or their percentage). As we already know, the Happy path would cover [30%-85%], so we need to list and to calculate the usage (ie. percentage of the coverage) of all the other variations.
Looking at this distribution would also be interesting. In the example above we have a distribution (Process Variation / usage) like this:
- Happy path: 36,74%
- Variation 1: 16,08%
- Variation 2: 13,64%
- Variation 3: 7,85%
- Variation 4: 5,87%
- Variation 5: 4,85%
- Variation 6: 3,58%
- Variation 7: 3,13%
- Variation 8: 2,90%
- Variation 9: 1,68%
- Variation 10: 0,84%
- Lower than 1%: Negligible or Process outliers.
The chart resulting from this distribution should look like a long tail chart. This kind of chart has a large number of values at the lower end (the “long tail”) and a small number of values at the higher end. In our case the biggest value (the first in the left below) is the Happy Path and could be almost alone to be so high.
Having a chart like this clearly means the data collected to rebuild the process looks consistent (for a statistics point of view) and this is the signal we may go further in the analysis.